Relational databases, which store tables of data where there are links between the tables, are used in many application domains such as accounting, product stock control, customer relationship management, product recommendation systems and others. Administrators of relational databases who control data entry to the database typically strive for accuracy and quality of data in the database. This is because the accuracy and quality will influence control of downstream systems, actions or other event which occur as a result of the contents of the database. However, accuracy and quality of relational databases are often reduced for many reasons including human error, limitations in opportunities and time to obtain data, limitations in equipment used to observe data and others. Users may be aware of missing data in databases. However, in many situations users may be unaware of errors and inaccuracy in data present in databases.
The embodiments described below are not limited to implementations which solve any or all of the disadvantages of known relational database management systems.